Problem difficulties could probably be done by redoing the almost completely unused rating system. Make it for the first post only, have an option allowing the thread's rating visible on the forum's listing of threads (maybe also one to make it completely invisible even in the topic), don't make ratings follow users. There was some discussion about this here, where the idea of only having a few people allowed to do ratings was brought up (how my name got on that tentative list I'll never know).

Will people actually use it?

From a moderation standpoint, I like these suggestions. A couple things I would like to see (though I've talked with Valentin about them already):

1. Allow one to merge the first post of a thread into another. Especially right after the AMC forum reopens, I often see duplicate threads. It would be nice to be able to completely merge them along with their first posts, which often contain important thoughts.

2. Fix the issues with the PDF download. I get one or two PMs about this each week, and I send out the same response each time.

I also would love to see that competition items from the resources sections can be downloaded as complete document, i.e. a .pdf-file with all available years instead of doing it year-wise for a particular competition.

Also... fix the typo in the HSB subtitle?

Just curious-- why isn't there a geology and geophysics forum? It looks like the one major branch of science you've left out, and from what I've seen, certain branches of it can be quite math-heavy (like igneous and metamorphic petrology, structural geology, and geophysics).

Hello,

From a technical standpoint, rewriting the whole site from scratch will almost surely be a bad idea. For several reasons:

- the code behind mathlinks has "grown hairs" throughout the years and it's easy to miss all those small improvements made when you're starting from scratch

- migrating the database to the new format will be a real pain, because the current database is being updated every moment

- you need to keep fixing bugs and do "critical" modifications to the current version of the site before you launch the new one. These have to be continuously ported to the new version of the site

- you need to have some sort of deadline and break your work into easily measureable pieces (when you start a project and say "we're very far from being done", in my head it translates to "we want to make these huge changes but have no idea if we can or when it could be ready")

- it would have a great impact on users (not all users are good at using computers). Users are adversely affected by huge changes - see the change from XP to Vista.

That's why I'm suggesting an incremental approach. Make a list of desired tasks (a lot of them should sound like "clean up the code in XX module"). Solve them one at a time, and deploy them to production. That way, the users won't feel much, but in a year or so, a lot can be changed. If you need more information on an incremental approach, you can read about how the Linux Kernel is being developed. And that's a piece of software with a huge impact!

From a user point of view, I also have a few observations:

- you waste (in my opinion) the most important area in the portal homepage by displaying static content (blog posts which are updated very rarely - and usually they're not "hot topics"). You can confirm this by installing a heatmap and seeing which parts of the homepage are most frequently clicked on. Normally, the center of the page should be the "hottest area". I suspect that in your homepage, this is not the case.

- when I go to the forum homepage, I'm totally lost in the hundreds of forums. I know I can hide them myself but it's a pretty tedious job. For myself, the following approach would work: if I'm logged in, display the first 5-10 forums I usually visit and have a button next to this list saying "show more" which will lead me to the complete list of forums. It's trickier for the users navigating anonymously - perhaps determine the general trends by using the data in Google Analytics and do the same thing - display the general trends. In any case, I think you should re-think the user interface and gradually alter it to be more easily usable!

Sorry for the long comment, blame it on Orl, he asked for my opinion

You identify a lot of the reasons we want to restructure the forum.

I'm not so interested in tiny things at this point -- when I say rebuild the forum, I mean a fairly large restructuring.

As for adding new forums, part of the point of this restructuring is to give users the ability to create new communities within AoPS. So, if people want to have a <whatever> forum, they can make one.

As for incremental changes, we'll make incremental changes on top of the larger switch. The community will still be basically PHPBB under the hood -- we're not rebuilding *everything* (although we will be converting to PHPBB3.0, which has much better moderation tools). We are rebuilding a great deal of the AoPS site, though -- it was built well before many items on the site even existed. And, as you note, the forum has clearly outgrown its current incarnation. We need a different way to navigate the many forums on the site.

I'm not worried about the huge changes issue -- users are mostly affected by huge changes when they are huge changes for the worse (XP -> Vista), not changes for the better (2000 -> XP). Besides, people will be able to use the site essentially as they do now. There will just be a lot more functionality over time. As for adding the functionality, we'll do that very incrementally after creating the core. Moreover, part of my hope is to make it much easier for new users to get started. Right now, the site is pretty overwhelming for a new user.

Typo fixed. That's just plain funny.

It's been there for at least a few months... I PMed both you and Valentin but never got a response.

I must've missed the PM; I generally take care of those things as soon as they're noted.

Back on topic, does anyone have any suggestions for how to organize information? For example, I probably move at least half of the posts that are placed in HSB into other forums because the posts are way too hard. How can we make it more clear to people where they should and should not post? (Or is this just impossible.) At the very least, the high school forums need to be rethought to some degree -- it really isn't clear where to post, so I can understand why people just post in the first forum they see. But how can we do a better job of organizing the forums to make more clear where they should post?

A couple of things...Maybe bad:

1) Probably too ambitious: For the difficulty ratings, can You make one forum where all the problems would go, and then a selected group of very active people would sort them? It wouldn't be necessary one forum, it could be on sub-forum inside all categories. But, I think it would fail, like the Solved Problems category failed.

2) Again problem ratings: DO NOT allow everybody to rate problems. Allow only active people who will judge fairly.

3) Olympiad section is very wide... I can see very hard problems and the next would be very easy. No idea here for now.

4) When I look at the Resources, I see lots of international competitions and don't know how hard the problems are. There should be ratings for each competition, so people could know what problems to do. I mean, TST-s in China and USA can't be compared to ones in small countries which don't have much potentials (I don't want to insult any country by writing its name here), but still the problems stand one next to another with no visible difference in difficulty.

5) As for the tehnical bugs, I have no comment as I am not so good with computer technology.

6) New users' problems: There can be hundreds of forums, but split them up in as many categories as you can.

7) I would like to see PDF problems with solutions. Someone can take the best appropriate solution in the forum (with appropriate, I mean that trig solution isn't the best for MATHCOUNTS) and just LaTeX it into PDF, and that would be much more helpful for users (at least me).

Hi,

I think that putting problems into categories is not a good idea. Because there are many problems which belong to two or more categories. That's why I'm suggesting tagging: tag a problem with tags from a predefined set (for example, one problem could get: combinatorics, Romanian TST, graph theory). Also, it would be a lot easier to browse for problems / solutions. Tagging works well, look at all this web 2.0 craziness.

Also, it's not feasible to ask users to estimate the difficulty of a problem and put it in the right category. Most of them do not have the necessary experience or pedagogic skills. I suggest an automated approach:
- give users a score (increases with number & quality of posts)
- let users give marks to problems, which will be weighted against their score. The oldest / most experienced users get to have a bigger impact

And what I meant to say about "small improvements" was: surely you customized PHPBB 2 *a lot*. All those customizations translate into features and fixed bugs that users get used to. They will miss them if you forget to make / can't make the same customizations to PHPBB 3.